Vol. XVIII 

 igoi 



I Clark, The Classi/icaiioti of Birds. ^ 7 ^ 



Auk and a Loon, it will at once appear that though the body 

 shape is very similar, the pterylosis is strikingly different, while 

 on the other hand, birds with very differently shaped bodies, some- 

 times have the pterylosis very much alike ; for example, a Goose 

 and Petrel, or (on the dorsal surface) a Loon and a Flamingo. 

 That the pattern of the pterylosis is very slightly modified by 

 changes in habit, is admirably shown in the Water-ouzels, where 

 the feather tracts are similar to those of the Thrushes, although 

 the habits of the bird have caused the development of a dense 

 coat of down over the whole body, a condition unique among the 

 Passeres. In the light of these facts it is clear that we cannot 

 assume that the pattern of the pterylosis is a recently acquired 

 and unstable character, and we must look for further evidence as 

 to its value. An examination of some Hummingbird embryos, 

 just before hatching, shows that the characteristic pterylosis is, 

 even then, clearly marked out. This would seem to indicate a 

 deep-seated character, for otherwise the feathers ought to appear 

 uniformly on the back and underneath, and only assume the charac- 

 teristic arrangement with the growth of the bird. A good illustration 

 of the fact that the line of development of a special form of ptery- 

 losis would appear in the embrj^olog}^ of the bird is afforded by 

 the Swallow, which, as is well-known, has a very characteristic 

 and peculiar saddle-shaped dorsal tract. In a large series of 

 young Eave Swallows {Petrochelidon fidva) from Jamaica, ranging 

 in age from embryos which are just assuming the form of a bird 

 up to those large enough to fly, I find that the dorsal tract as first 

 marked out is much more like that of Swifts, than Hke the adult 

 Swallow ; that is, it is first a broad, dorsal patch with a small cen- 

 tral apterium. Later on, it begins to assume more the condition 

 characteristic of the adult, but that condition is not fully attained 

 until the bird is able to fly. It seems to me that the young Hum- 

 mingbirds, and the young Swallows together, show that the pat- 

 tern of the pterylosis follows the same laws of heredity as other 

 characteristics, and the later modifications of the pattern appear 

 later in the development of the individual. Thus the Humming- 

 birds, being nearer the ancestral form in the pattern of the ptery- 

 losis, show that pattern from the start, while the Swallows, being 

 more specialized, simply pass through that stage in the develop- 



